Browse Reports

We study the problem of scheduling a set of hard-real time constrained tasks on one processor, each of which is defined by a tuple (r,c,p,d), where r is its release time, c computation time, p period, and d deadline. The objective is to decide whether all tasks can meet their deadlines at run time on one processor. If so, we say that the task set is scheduleable. We prove two sufficient conditions for a (restricted) periodic task set to be scheduleable. The two conditions can be applied to both preemptive and non-preemptive scheduling, in sharp contrast to earlier results. We also prove a necessary condition for a periodic task set to be scheduleable under any scheduling algorithm. Our second sufficient condition is most general with respect to scheduling algorithms which do not idle the processor as long as there are tasks ready to execute. If a periodic task set is scheduleable under any such scheduling algorithm, it is also scheduleable under our second sufficient condition. We then present a method for transforming a sporadic task to an equivalent periodic task. We show that the transformation method is optimal with respect to non-preemptive scheduling. With this method, all results on scheduling periodic task sets can be applied to sets of both periodic and sporadic tasks. Key words: real-time, hard real-time, real-time control, scheduling, real-time scheduling, hard real-time scheduling.